Medical and hospital care for the 60,000 residents of Rochester, Minnesota, is almost entirely provided by Mayo Clinic and the Olmsted Medical Group and their affiliated hospitals. The records of these institutions and other providers have been assembled into a single diagnostic retrieval system and provide, for several decades, a unique resource for population-based epidemiologic studies of serious chronic diseases. Previous studies of coronary heart disease (CHD) in the Rochester population established an incidence cohort for 1950-82. This data base has provided incidence rates for angina pectoris (AP), myocardial infarction (MI), and sudden unexpected death (SUD). Follow-up for the local population is exceptionally good and has provided optimal estimates of long-term survivorship, infarction rates in the angina cohort, and reinfarction rates in the initial MI group for both transmural and subendocardial infarctions. It is proposed to extend recent incidence and trend studies through 1988 to determine whether there is a continuation of trends such as increasing MI incidence in females and a recent decrease in males. There has been a decrease over the same time period in SUD incidence rates and a sharp fall in the 1970s in the case fatality rate for MI. In the AP cohort, MI rates have decreased and the long-term survivorship has improved; while in the MI cohort, there has been no drop in the reinfarction rates and long-term survivorship has not improved. The incident CHD cases will provide a basis for monitoring a variety of invasive and noninvasive diagnostic procedures and lead to sound planning for future needs for therapeutic procedures such as ventricular assist devices, artificial hearts, laser catheters for coronary stenosis, etc. Comparison of referral and local practices will help identify the effect of referral bias on patient characteristics and prognosis in CHD reported in the literature. Also, the effect on CHD incidence rates following the introduction of Diagnostic Related Groups (DRGs) and the mandatory diagnostic listing for application of the DRG system will be assessed. There are few, if any, long-term U.S. data on the occurrence of AP, MI or SUD as the initial manifestation of CHD other than Framingham and the Rochester data. These data form a crucial complement to the national mortality data.